FIGURE 4-1 Physician specialties providing care for generalist-dominant diagnosis cluster, NAMCS 1989–1990 (all specialties are listed that accounted for ≥ 10% of visits for that condition). All relative standard errors are ≤ 30%. AI = allergy and immunology; Internal Medicine (IM) = general internal medicine; OBG = obstetrics and gynecology; ORS = orthopedic surgery; OTO = otolaryngology; Pediatrics = general pediatrics. Unlabeled cells represent other specialties. SOURCE: Rosenblatt et al., 1995.

To measure accurately the effects of primary care, data systems and research methods must be able to reflect the co-occurrence of health problems and the longer time frames needed to evaluate the integrative functions of primary care—comprehensiveness, continuity, and coordination. However, in primary care, recording the content of care using standard coding systems such as the International Classification of Diseases or the Diagnostic and Statistical Manual (for mental health) is inadequate because of the complex interplay among diagnostic categories and other clinical or social problems a patient may often simultaneously experience. Current U.S. databases, such as those provided by NAMCS, are by design cross-sectional analyses of single visits and thus cannot capture episode information.

These databases compile diagnostic data from the point of view of visits to clinicians reflecting the practitioner perspective rather than the perspective of the